1Password targets AI token spend as enterprise budget crisis looms
1Password is extending its SaaS governance platform to track AI token consumption, betting that unpredictable model pricing will spawn a new enterprise cost-management category.
1Password on Tuesday launched AI Spend and Consumption Management, a tool within its SaaS Manager platform that provides finance and IT teams with a real-time view of enterprise spending on large language models. The product connects directly to the admin APIs of Anthropic, Cursor, and OpenAI to aggregate daily token-level data into a single dashboard. It is currently in public preview, with broad availability planned for fall 2026, and is included at no additional cost for existing customers.
The release targets a structural breakdown in corporate budgeting. Traditional software is sold on predictable per-seat contracts, but AI services charge per token, where costs fluctuate based on the model used and task complexity. Autonomous AI agents executing multi-step workflows can drain prepaid budgets in hours, often without triggering alerts until the invoice arrives.
Greg Henry, 1Password's chief financial officer, compared the current AI spending chaos to the early days of cloud infrastructure. "Consumption-based pricing isn't new," he said. "We saw it arrive with cloud infrastructure, and it took years to build the tools and disciplines to manage it. AI is the next version of that shift." That parallel matters for investors: the inability to track early cloud spend birthed the multi-billion-dollar FinOps ecosystem, including companies like CloudHealth and Apptio.
The scale of the impending budgetary pressure supports 1Password's strategic pivot away from consumer password management. Goldman Sachs has projected that token consumption from AI agents alone will grow 24 times by 2030. As these autonomous systems handle increasingly complex tasks, the volume of API calls will outpace the capacity of manual tracking.
The system normalizes disparate vendor billing formats and disaggregates costs by team, user, and model. "You can't enforce what you can't see," Henry said, explaining why the current product focuses on visibility, issuing threshold alerts rather than automatically cutting off agent-driven spending. The company is "actively evaluating" automatic enforcement for future releases.
The tooling gap is forcing a structural shift in corporate decision-making. Because the price disparity between different AI models has become significant, the choice of which model an engineering team uses is now a financial decision. "AI spend can't be treated as a finance-only or IT-only problem," Henry said, noting that chief financial officers are being pulled into product discussions in unprecedented ways.
Early adopters report concrete benefits from the added visibility. "Forecasting tools for AI consumption and spend was one of our biggest gaps in planning," said Steve May, director of IT at ServiceTrade. He noted that the visibility has already "prevented overages that would have cost far more to fix after the fact."
